‘Social distancing” remains the operative term as we try to slow the spread of this harrowing COVID-19 pandemic. To date, 41 states have issued stay-at-home directives, on top of recommended curfews, discouraged social gatherings, and other forms of self-isolation. As necessary as these measures appear to be, potentially saving millions of lives, they also cause considerable economic damage and social interruption. Some — President Trump included — are ready at least to discuss a return to normal life.
Determining how and when to safely resume the routine of work, worship, travel, and recreation is uncharted territory. Since reliable infection data are scarce and rapid response is necessary, public officials are turning to a novel source of data: our smartphones. Routine, anonymous location data could accelerate a return to normalcy.
The suddenness of the economic shock is unprecedented. Over 6.6 million Americans have filed for unemployment benefits so far, and some estimates have the unemployment rate at close to 13 percent, the highest level since the Great Depression. Further, J.P. Morgan economists estimate that GDP could be down as much as 13 percent in the second quarter of 2020. While social distancing is necessary, it’s costly.
Meanwhile, medical researchers at the University of Washington currently estimate that the use of hospital resources will peak and ebb in most of the country in two or three weeks, around mid or late April.
For policymakers to begin easing social-distancing restrictions and allowing people to return to their day-to-day lives, they first need to know when and where the infection risk is low. In addition, public-health officials need to know which areas are in the worst shape, so they can reallocate scarce medical resources to the places that need them most.
One way to do this is by using aggregated, anonymized user location data obtained from smartphones and apps to assess social distancing behaviors in different parts of the country. The goal is not to examine the actions of specific people, but to analyze trends by combining significant amounts of unidentifiable information from a large body of individuals.
Most American adults carry an always-on geolocation device — their smartphone — on their person most of the time. Many popular smartphone apps passively transmit user location to app companies at all hours of the day, via cellular, GPS, WiFi, or Bluetooth technology. These millions of data points are used by app companies to target users with local ads or to improve proximity-based services such as dating apps, mapping apps, or recommended restaurants.
The use of smartphone location data always poses potential privacy risks, so policymakers, health officials, and tech companies should agree on narrow uses of the data. They can do this by ensuring that the data are anonymized, limiting their use to the duration of the crisis, and making it only available to public-health officials (such as those at the CDC) rather than law-enforcement agencies.
Smartphone location data are already being used by officials in Austria, South Korea, Taiwan, and Italy. Similar efforts are starting to take place within the United States. Just this past week, Google released aggregated, anonymized location data in an effort to make policymakers and the public aware of county-wide mobility trends. Facebook is sharing similar data with health officials. Assessing the specific needs of those areas is the next step.
President Trump recently encouraged governors to categorize counties based on their risk level. Leaders can use location data in combination with infection and hospitalization rates to classify areas as high-risk, moderate-risk, and low-risk.
For example, areas with high levels of infection and hospitalization — and where location data shows higher-than-expected resident mobility — should be categorized as “high-risk.” Areas with few infections and strong social distancing are “low-risk.” “Moderate-risk” areas would fall somewhere in between, or be strong in one metric but not the other.
With a better sense of where the risk is high and where it is low, policymakers could then do three important things. First, they could reallocate masks, ventilators, medical labor, and other high-value resources to the higher-risk areas. Second, once testing and masks become more widely available, they could push these supplies to lower-risk areas. Asymptomatic carriers of the virus would be better aware of the risk they pose to others and more likely to self-isolate. Officials could slowly begin relaxing some social distancing enforcement policies. Third, they would have an opportunity to better assess and improve the efficacy of stay-at-home orders and education by comparing the outcomes in different states and cities.
Relaxing social-distancing policies will need to be done with caution, accounting for each location’s unique conditions, such as transportation hubs, common travel destinations, and proximity to other, higher-risk areas.
While we may still be months away from a national return to normalcy, it’s not too early to start planning for better times. Decision-makers need quantitative, ascertainable benchmarks to make such high-stakes decisions successfully. Smartphone location data have their limits. But they will provide useful information at a time when many public-health officials, doctors, and residents feel like we’re flying blind.
Brent Skorup is a senior research fellow with the Mercatus Center at George Mason University. Trace Mitchell is a Mercatus research associate. They are coauthors of the new policy brief “Aggregated Smartphone Location Data to Assist in Response to Pandemic.”